Colorectal cancer (CRC), the second most common cancer killer in the U.S., is a cause of substantial death and suffering. Adequate staging of this disease is essential; node status (positive or negative) determines prognosis and treatment. Patients with positive nodes are offered adjuvant chemotherapy. Additionally, patients who have more nodes evaluated have a better prognosis than those who have few. Despite the importance, and clear criteria (at least 12 nodes), most patients (56%) in the U.S. do not undergo adequate node evaluation and thus node staging must be improved. Our study proposes to determine provider, patient, and tumor factors that influence node evaluation, and will investigate the underlying association between node staging and survival. The study will provide critical information to develop targeted strategies so that the quality of staging can be improved for all CRC patients. In addition, through our research we will gain a better understanding of the mechanism underlying the survival benefit seen in patients who have an adequate node evaluation, and may provide significant insights into underlying tumor-host biology. We will use SEER-Medicare linked data to identify patients (almost 40,000) treated for nonmetastatic CRC over an 8-year period. We will determine the number of nodes evaluated for each patient and will determine the effect of provider (hospital, pathologist, and surgeon), patient (age, race, socioeconomic status, comorbidities), and tumor factors (grade, T stage) on node number. Important predictors of node number will be potential targets for intervention strategies. Such targeted strategies are necessary to ensure that all CRC patients in the U.S. have accurate staging, and that all node-positive patients who could benefit from adjuvant therapy are identified. The study will then evaluate the relationship between the number of nodes evaluated and outcome. We will test the assumption that the improvement in CRC patient survival with higher node number is due to upstaging (improved prognosis by stage because of more accurate categorization of patients, with no true improvement at the individual level), at either the patient or provider level. We will then perform survival analysis, using a Cox proportional hazards model to determine if number of nodes evaluated remains an important determinant of CRC patient survival after adjusting for potential confounders including provider, patient, tumor, and treatment factors; if so, node number, a reflection of host-tumor immunologic interaction, may be an independent prognostic factor. This information would have treatment implications and would necessitate further research in this area. Most CRC patient in the U.S. do not have an adequate node evaluation. This research will identify factors that affect the adequacy of lymph node staging in CRC so that improvements can be made. [unreadable] [unreadable]